Feeling Nervous? - March 20, 2012

[Pascal]

Thanks, and I agree. I'll take this one step further, and maybe Pascal can comment. I'll go out on a limb and give him something specific to shoot down if I misstate / misunderstand what we are all doing here.

Looking at where we are in MF in context, the 20DMF reminded me that we are still < 0, i.e. in negative territory, and the MF trend is clearly down. In the last 5 days, of course, the MF has been clearly up, as seen by the RT charts. So compared with 5 days ago MF is up - big money moving in. Over the last 20, however, big money has moved out.

What's noise, and what's the trend? There's the question for model. In other words, what - and how much - does a clear 5 day positive change in MF mean within the context of (1) absolute negative MF (e.g. negative number) and (2) a longer term (e.g. 20D trend/change) negative MF.

Obviously there's a profound implication for the value of the RT vs. EOD model in the answer to that question. The more the long term matters, the more the short term, and therefore RT model, becomes irrelevant. The more the (immediately detected) short term changes drive future price changes, the more the RT model becomes invaluable.

I can't believe that volume (MF) doesn't drive price. That's an axiom of technical trading. So it can't be that we are watching money move in in volume with no price increase. I just won't believe that. Either (1) we aren't measuring volume correctly (and I'll trust Pascal that that isn't true), or (2) extremely short term volume (whatever that period is - 1 minute? 5 days? 10 days?) is noise.

Djones,


Your observations are totally correct in the sense that MF and EV are similar animals: they are useless if you do not use them in a "value" context. When I studied EV, I observed that the three days trend of EV usually moves in opposite direction to the three days price trend and hence, EV is a misleading indicator if you use it out of context - I wrote about this in the VIT book.

However, EV and more specifically LER (the ratio of LEV to the total volume) offer good opportunities when they are used with a specific "value" detection method. AB is a good value detection method. Support/resistance are also good methods. Another aspect of the EV tool calculated on a single stock is that the price of that stock will also be influenced by the price of the stocks in the same sector, simply because funds diversify through a sector and many individual traders now use ETFs, which is similar to spreading money across one sector. This is the reason why I always advocate the use of EV in the context of a sector's buy signal for example.

In the case of the GDX MF, we have the advantage of measuring EV for all the stocks of the ETF, transforming this measure into a weighted MF measure. This means that this measure is stronger than an EV measure on a single stock.

But still, it is important to use this MF "in context." The GDX MF model uses its OB/OS levels - and other levels - as a context. When the MF hits the OS level and moves back above it, then it simply means that large players have been "flushing things out" and buying. So the model (RT model here) buys.

The EOD model however looks at the same context only at the close. It is therefore slower to react, which means that it is better to trade EOD signals with non-leveraged instruments.

The RT MF shows a different aspect of the trading context: the MF/Price relation. It allows to see in RT how the big money reacts to price. For example, when we have a down-gap, is this perceived as a buying opportunity or as a sign that the sector is largely oversold and that it is better to move out.



Pascal

PS: By the way, The Figure below also shows what the right handle is used for (Bob's question)